September 1967
783
Totally Synthetic Steroid Hormones. XV1I.l Further Studies on the Synthesis of dl-WMethylandrostane and dl-18-Methylpregnane Derivatives RICHARD
REES,D. P. STRIKE, BND HERCHEL SMITH
Keaeuith und Deueloprr~entDzvzszons, TYyeth Luboratoi ies ItLc., Budnor, Ptrina~liuniuL
Receiued M a y 1 , 1967 tll-l8-~Iet~hyliestosterone( X ) and dl-18-methylprogesterone ( X X I I ) have been synt'hesized through the Simmons-Smith methylenation of dl-li@-acetoxy-18-methylestr-3(lO)-en-3p-ol (IT). dl-18-?\Iethylandrost-4-en-3,lidione (XT') has been synthesized through interacting d l - 1 7 ~ - a c e t o x ~ - ~ c u , l O ~ - e p o x y - l 8 - m e t ~ h ~ l e s t r(aXnI-I~)~ - o l with CHJLlgBr. The pure dl-lip-acetoxy-18-met~hylestr-S(10)-en-3a- and -3p-ols 111 and 11-have been prepared from dl-lip-acetoxy-l8-methylestr-S( lO)-en-3-one (11)by a novel seqiience involving epoxidation, NaBHa reducI ioii, separation into the correspoiiding piire SLY,10or-epoxy-3cu-nl XI1 .i,lOp-epoxv-:lp-ol XITI. arid deepoxidation. dl-li~-AcetoxS.-18-methylestra-S~ 10),0( ll)-dieii-:3a-ol i X X \ - I ) has beeir selectively methylenat,ed at)t,he 3,lO positioii by the Siminoiis-Smith reagelit aiid the pi.odiic.1 has heel1 lraiisformed t ( i H hteroid foriniilated H S dl-lip-acetox~-!~~-chl~ror1~ethylestr-4-ei1-~~-~1ie ( X X S ) . The 1)rogestatiotial niid aiitie>tiwgeiiic. activities of rl(-l8-meth~lpl.ogesteroiieX X I I are reported aiid compared with the correspoiidiiiy activities of progesteroiie, l!)-norprogesteroiie, aiid tll-18-methyl-lQ-norp~ogesterorie.
The biologically important properties which we recently for a series of totally synthetic dl-18-methylestrane (13P-ethylgonane) derivatives necessitated the preparation of appropriate 10P-methyl analogs. The preceding paper' described the total synthesis of several of such analogs through the introduction of a single carbon substituent at the 10 position of the dl-18-methylestrane nucleus. We nou- report syntheses using the same principle which provide alternate routes to dl-18-methyltestosterone, dl-18methylprogesterone, and related substances.
dl-18-Methyltestosterone and dl-18-Methylandrost4-ene-3,17-dione.-Because 3-oxygenated 18-methyllestrane derivatives were readily available from our (earlier J V O we ~ ~chose ~ t o prepare dl-18-methylandrost4-en-3-ones through the Simmons-Smith methylenation'o of appropriate lS-methylestr-j(lO)-en-3-ols. Sotably, testosterone has been obtained through a Simmons-Smit h reaction with the 17P-2'-tetrahydropyratiyloxyestr-5(lO)-en-X~-ol present as a minor component in admixture with the Corresponding 3aepimer, and dl-6cr-androst-4-ene-3,17-dione has been formed by related experiments in the dZ-Sa-estrane series.I2 Since the Simmons-Smith methylenation of homoallylic alcohols is stereochemically controlled by the configuration of the hydroxyl, 11,13 dl-18-methylestrI : D. P. Strike. I). Herlist, a n d H . Smitll, J . .lIu/. Chcm., 10, 446 (1967). (2) Postal address: P.O. llox 82!18, Pliiladeliiliia, P a . 1Yl01. ( 3 ) ( a ) H. Smith, et nl., A ' z p e r i e v f i / i , 19, 3'31 ( I M 3 ) ; (11) . I . Chrni. , S U L 4472 (1964); fc) G . H . Douglas, G . C. Iluzliy, .Jr., C. R. \Valk. anrl H dmitli, Tetrahedron, 2 2 , 1019 (1966). Edgren, H. S m i t h , L). L.Peterson, a n d I). L. C a r t e r , Steroi,is. 2 , 319 (1968). ( 5 ) R. .i. Edgren, H. S m i t h G . .'Hughes, i . L. L. Smith, and G . Greenspan, i 5 i d . , 2 , 731 (1963). (6) R . 11. Tomarelli a n d F. JV. Bernhard, ibid., 4 , 4,5l (1964). ( 7 ) R. A . Edgren a n d H. Smith, "Hormonal Steroids, Riochemistry, Pharmacology, and Therapeutics, Proceedings of t h e First International Congress on Hormonal Steroids," Vol. 2 , Academic Prevr Snc,, New York, K . T., 1965, p 161. ( 8 ) R . A . Edgren, D. L. Peterson, R. C . .Jones, C. L. N a g r a , 1%.Stnitti, anrl G . A. Hughes, Recent P r o g r . Hormone K e s . . 22, 805 (1966). ( 9 ) C . .I. Hughes, T . Y. J e n , a n d H. Rmith, S t r i o i d s , 2 , 9 4 i (1966). (10) I t . E. Simmons aiid R. 13. Smith, ./. .4m. Chern. S o c . , 80, 5323 (1998); 81,4256 (1959); H. E. Simmons, E. P . Ulanchard, a n d R. L). S m i t h , ibid., 86, 1347 (1964). ( 1 1 ) Et. Uiriaii: and .I. U. Cross, ihid.,87, -162'3 (191%). (12) .I. ,J. Birch i t i d 0. S. It. Sul~ba.Itao, J . Chern. Soc., ;I39 (lY6:).
-5 (10)-en-3P-ols were required. Reduction of the acetate 11, of the corresponding alcohol with sodium borohydride in methanol gave a difficultly separable mixture of the corresponding 3-01s judged by thin layer chromatography t'o contain the epimers in a 4:1 ratio. By analogy with previous findings on the reduction of 176-acetoxy- and -propionoxyest'r-5(lO)-en3-ones under similar conditions,'j the major component' is expected to be the unwant'ed 3a-01 111. Aft>erchromatography on silica gel, the mixture contained approximately equal amounts of the epimers and was treated with the Simmons-Smith reagent lo (prepared from an appropriately activated zinc-copper coupleI6) to afford a crystalline product' apparently containing the cyclosteroid VI1 and unreacted 111. This mixture, after oxidation n-ith chromium trioxide in pyridine, '' readily yielded pure VI11 and the latter, on keeping at' room temperature in CHCl, saturated with HCl arid reacylat'ing the partially hydrolyzed product', gave dl18-methyltestosterone acetate (IX) which was identical with a sample prepared through the addition of hydrogen cyanide to 17~-hydroxy-lS-methylestr-5(10)en-4-0ne.l A similar process starting from pure IV (below) and involving a final saponification gave dZ-18methylt'estosterone (X), identical n-it'h the previously described subst,ance.l Having established the feasibility of the rout'e from IV to IX, we next desired to increase the yield of IV. The use of different reagents for reducing I1 offered little hope for improvement since the corresponding lcetol I with Li-41H4, lithium tri-t-hutoxyaluminum hydride, or Adam's catalyst in ethanol gave mixtures of 3-h>-droxyepimers corresporid1,3114
(I:+) S. \\'inatein, .J. Sunnerliercr, and L. IIeVries. .I. Am. Ctiem. S'oc., 81, S52:3 (195'3): Y. \\-instein and J . Sonnerlierg, ~Oiii..83, 3235 (1961): E. J . Cores and R. L. D a n s o n , t b i d . , 8 6 , l i 8 2 (1963); \\'. G. Dauben a n d G . H. Berezin, i b i d . . 8 6 , 168 (1968); I V . G. Dauben and h. C. .Ishcraft, ibid., 8 6 , 367X (1967); P. Radlick and S.IVinstein. i b i d , , 86, 1866 (1964). (14) This anrl other racemic steroids desrrilied in this paper are depicted Iiy t h e enantiomorphs having the 13-alkyl group in t h e 0 configuration. Enantiomorphic steroids with tlie same absolute configuration a s t h e naturally occurring series are given tile prefixd in accord with t h e convention proposed b y L. F. Fieser a n d 31. Fieeer, "Steroids," Reintiold Publishing Corp.. N e w Y o r k , K ,Y . , 1959, p 336. ( 1 5 ) ( a ) d . G . l.evine, N. 11. E i i < l \ - ,and l i . (', Fartl,iiig, Tutritiirdru,, L u t t s r a , 1 5 l i (1968); (b) J . Org. Chem., 31, 3895 (1966). (16) E.LeGofi, ibid., 29, 2048 11864). ( 1 7 ) G.E. . i r t l i , G.1. I'uus, 11, 31. Lukeb, 1 . M. Iloliiiison, \V. 1'. J o l ~ n s , AI, k'eurer, and L.11. Sarett, J . A n i . L'iism. S u c . , 76, 1715 (1954).
Scptcnibcr 1967
HO
O I, R = H 11, R = COCH,
p H
111,3a-OH;R = COCHB IV, 38-OH;R = COCH, V, 3a-OH; R = H VI, 3 p O H R = H
o&
H
H
H
HO VI1
VI11
7s3
XVII
TOTALLY SYNTHETIC STEILOIDS.
Hydroboration of XX follou ed by H202 oxidation gave the crude alcohol XXI to which the 200-hydroxy configuration can be assigned, assuming the usual 0face attack of diborane and oxidation of the resulting borine without inversion of configuration. Jones' oxidation followed by treatment a t room temperature with CHCla-HCl transformed XXI to the required dl-ISmethylprogesterone X X I I in 13% over-all yield from VIII. The 17P-acetyl configuration in X X I I is confirmed by the pmr spectrum which displays the 18amethyl protons as a triplet centered a t 6 0.70 in agreement with the value 6 0.67 found for ls-methylprogesteroneZ7and dl-18-methyl-19-norprogester0ne.~ As with dl-l8-methyl-19-n0rprogesterone,~ treatment with methanolic IZOH equilibrates the 170-acetyl group in X X I I with formation of a mixture shown by pmr spectroscopy (cf. ref 9) to coritain the 170 and 170 epimers in a ratio of ca. 1:3 .
dH
pLJB
IX,R = COCH3 X,R=H
e
XI
e 0
pJp ,yp p & c XX
XI X
\
HO
A
I
, COCH,
CHOHCH,
H
/
H
H
OH
XII, 3a-OH5a,10a.epoxide XIII, 38-OH;5410j3-epoxide
XIV
XXI
XXII
Simmons-Smith Methylenation of dl-l7p-Acetoxy18-methylestra-5(10),9(ll)-dien-30-and -%p-ols.--We wished to determine whether the Simmons-Smith methylenation of a gona-5(10) ,9(1l)-dien-3-01 would be 0 selectively directed to the 5(10) double bond by the 3-hydroxyl group. The alcohols XXVI and XXVII XV x VI required for this study mere prepared as a crystalline mixture by reducing the dienone XXV with SaBH4, XXV being obtained from I1 by a standard routez8 involving bromination-dehydrobromination, conversion of the resulting dienone X X I I I to the enamine XXIV, and formic acid hydrolysis. Simmons-Smith methylenation of the mixture of XXVI and XXVII a t 90" XVII XVIII under pressurell afforded, after chromatography, two products in yields of 24 and 8%, respectively. The pmr pyridine,I7 to the oxoketal XIX, which was converted spectra of both substances displayed geminal cycloby ethylidenetriphenylphosphorane to the pregnene propyl and vinylic proton signals in the 6 0.5-0.70 and XX. The latter is assigned a configuration in which 3.40-5.75 regions, respectively, thereby confirming the C20-hydrogenis anti to the main bulk of the steroid methylenation at the 5(10) rather than the 9(11) posinucleus, on the basis of its pmr spectrum which distions. Catalytic hydrogenation of the mixture of plays the Czo-proton as a quartet centered at 6 5.18, XXVI and XXT'II gave an almost quantitative yield and the C21-protorisas a doublet a t 6 1.67. In a related of a product resulting from saturation of the 9(11) series of 17-ethylidene-18-methylestranederivatives double bonds and containing, from thin layer chrohaving the same geometrical configuration, these promatography, ca. SO% of 111 and 20yo of IV. It theretons have been observed in the ranges 6 5.22-5.25 and fore follows that the major methylenation product is 1.73-1.78, respectively, whereas in the corresponding derived from the major component XXVI, and, asgeometrical isomers in which the C*o-proton is syn to suming stereochemical control of the introduction of the the main bulk of the steroid nucleus, they lie within the h j r the 3-hj.tlros\-l groiip, js formulated rarigeb 6 4.99-3.09 :tiid I.LS-l.Gl, r e s p e ~ t i v e l y . ~ *Cl9-niet2ij.leiie ~~ (26) B. Gadshy a n d 1. 13 .\. J a n s e n ( J o h n W5etl1 a n d Brother Ltd., laplow, England) a n d L). P Stiike, R.P. Stein. and H. SInltll, unpublished
,>
\I ed. Farther elsboratiou of the resiilting li-alkyl derivatives was carried orit by standard procediires. Of particiilsr interest is the synthesis of the li-methyl and li-ethyl derivatives of li-deos~-~)a-flut,roprediiisoloiie 21-ttcetiite and of 17-deoxy17-methyldesamethasone 21-acetate.
In this paper we report t8hesynthesis of a variety of l'ia-methyl and ethyl derivatives of the corticoid type. Previous reports'* from this laboratory have described a useful procedure for the preparation of 17-alkylpregnan-20-ones, which involves the alkylation of an intermediate 17-enolate anion developed by treatment of a 16-pregnen-2O-one with a solution of Li, Ca, or Ra in liquid ammonia. The application of this method to the synthesis of a series of orally effective 17-alkylprogesterone derivatives has also been described.',* 1."- the present study we have utilized this method and also have developed a convenient process for the simultaneous introduction of met'hyl groups a,t the 16a: and 17a positions involving Grignard 1,4 a,ddition to a A1620-ketone followed by methylation of the intermediate enolate anion. Some years prior to this investigation Engel described the synthesis of 17-methylcorticosterone3 and its 11f l ) (a) 31. J . \\-eiss. R. E. Srhaiih, G . R. Allen, J r . , J. F. Poletto, C. I'idarks. R. 1.C'onrotv. ani1 C . .J. Coscia. T e t r n i r d r o n , 20, H R i (1964); ( b ) w? also H. I)wIienglii, C . RPvein, a n d R. Baildry, J . .Mri/. C i i r m . . 6 , :301 (I'Jii8). ( 2 ) AI. .J. \Veisn, It. I,:, Scliaiih. .I. 1:. l'oli-ttu, (;. I f . : \ l l ~ r ! , .Jr.. : I I ! ~(', l'iilwks, S t r r o i d s , 1, 608 (19tk3).
dehydro d e r i v a t i ~ e . ~From the results reported at that time it appeared that replacement of the hydroxy group at CI7with a methyl group results in a decrease in glucocorticoid and antiinflammatory activity as measured by liver glycogen and local granuloma assays, respectively. On the other hand, the possible effect of this modification on mineralocorticoid activity aroused our interest since it appeared that, with regard to this important parameter of biological activity, there was a qualitative difference between 17methyl-1 1-dehydrocorticosterone acetate and cortisone a ~ e t a t e . ~We have been unable to find any further reports concerning this preliminary observation nor, to our knowledge, has there been an application of the possibilities thus raised to a Sa-fluoro-substituted corticoid. With the hope that 17-alkylation in the corticoid series might in fact result in a favorable separatio:i of certain of the parameters of corticoid activity arid encouraged by our own observations? in the progesterone series that at least in some instuncc:, ly-ethyl ;mtl R. Engel, Con J C h e m , 86, 131 (1967). c7. R. F,ng?l .r 4 ,,I C ' h a n ~ ,hL , 78, 4727 (1956).
(3) C (4)
